William Liang personally invested nearly $3 billion to lead the first round of financing: DeepSeek's valuation soared to $51.5 billion, V4.1 to be released in June

According to Dongcha Beating's monitoring, DeepSeek's first-round funding saw a significant change. The Information revealed that the founder, Liang Wenfeng, plans to personally invest up to 20 billion Chinese Yuan, subscribing to about 40% of this round's shares. This means he will replace the previously rumored state-owned giant fund as the largest contributor to this $7.35 billion (50 billion Chinese Yuan) massive financing, with the China National Integrated Circuit Industry Investment Fund expected to fall to second place. The post-investment valuation for this round has also increased from the rumored $45 billion to over $51.5 billion.

With a clear capital structure, DeepSeek's commercialization strategy is now defined:

1. Core Update V4.1: Set to be released in June, with a focus on strengthening enterprise tools, natively supporting MCP (Model Context Protocol), and adding image and audio comprehension capabilities (generation side remains text-only).

2. Compensation and Option Defense Line: The new funds will be used to expand computing power, increase employee compensation, and issue stock options. This is a direct response to the reality of tech giants poaching talents—DeepSeek's V3 core contributor, Luo Fuli, has moved to Xiaomi to lead the AI department, while another core researcher, Guo Daya, was poached by ByteDance with a high salary.

Liang Wenfeng has shifted from a "symbolic follow-on investment" to "writing the largest check," thereby maximizing his control over the company. However, faced with soaring computing power costs and the loss of key talents, DeepSeek will eventually have to abandon the "no revenue pressure" pure research lab model and address the self-sufficiency issue.